The use of large scale computing resources has revolutionized the world of computing. Small, lightweight, low power mobile computing devices may rely on the use of the large scale computing resources to perform heavy duty processing, thereby minimizing the amount of power expended at the mobile computing device. In addition, large scale computing resources may provide enormous databases and store exponential amounts of accessible information that enable our modern technical infrastructure to operate.
The operation of the large scale computing devices typically involves operating thousands of servers in large server farms, along with the network switches, routers and other devices used to enable communication within the server farm and with the end users of the server farm. Each server farm may support thousands or even millions of end users. Accordingly, it is important that each server farm is highly available and has minimal down time.
To maintain a high level of reliability within a server farm, a great amount of redundancy is often employed. In addition to backup servers that may be employed in case of the failure of another server, backup power is also used. The use of backup generators and uninterruptable power supplies (UPS) may be used to reduce the inherent instability of power received from power stations. Integrated sensors in server farms may also be used to monitor power distribution to equipment within the server farm.
However, changes in phase, voltage, and current that intermittently occur in power received from power stations may reduce the reliability of the electronic equipment used in server farms. The use of integrated sensors in each of thousands of devices operating at the server farm may be expensive and cause additional points of failure, and integrated sensors may report on a limited number of power problems. The use of integrated sensors may actually make it difficult to monitor the power throughout the server farm and decrease the overall reliability of the server farm.